Latte & Learn: Exploring Energy-Efficient Architectures for LLMs

​TOPIC: Exploring Energy-Efficient Architectures for LLMs

As part of our weekly research paper study group, we are focusing on exploring energy-efficient architectures for LLMs. This week, we will be discussing the DynamoLLM Paper published by Microsoft Research.

We will explore how to optimize LLM inference clusters for both performance and energy efficiency. This paper provides an in-depth look at how DynamoLLM designs intelligent inference clusters to balance high-performance serving with minimal energy consumption.

The Deep-Tech Community:
We are building the 𝐛𝐞𝐬𝐭 𝐀𝐈/𝐌𝐋 𝐝𝐞𝐞𝐩-𝐭𝐞𝐜𝐡 𝐜𝐨𝐦𝐦𝐮𝐧𝐢𝐭𝐲 in Silicon Valley and around the world.

Along with regular meetup events in Palo Alto, SF, we organize invite-only 𝗱𝗲𝗲𝗽-𝘁𝗲𝗰𝗵 𝗶𝗻𝘁𝗶𝗺𝗮𝘁𝗲 𝗱𝗶𝗻𝗻𝗲𝗿𝘀, invite-only 𝗱𝗲𝗲𝗽-𝘁𝗲𝗰𝗵 mixers, a 𝘄𝗲𝗲𝗸𝗲𝗻𝗱 𝗿𝗲𝘀𝗲𝗮𝗿𝗰𝗵 𝗽𝗮𝗽𝗲𝗿 𝘀𝘁𝘂𝗱𝘆 𝗴𝗿𝗼𝘂𝗽 and 𝗱𝗲𝗲𝗽-𝘁𝗲𝗰𝗵 𝗹𝗲𝗮𝗱𝗲𝗿𝘀𝗵𝗶𝗽 𝗴𝗿𝗼𝘂𝗽.

Our goal is to bring together AI/ML deep-tech researchers from both academic and industry backgrounds, as well as VCs interested in the deep-tech domain. This event fosters collaboration and networking among these diverse groups, facilitating the exchange of ideas and the formation of new partnerships.

​​​​Who Joins Our Crowd

​​​​Our events attract a prestigious crowd, including:

• ​​​​Researchers from renowned institutions such as Stanford, UC Berkeley, and other prestigious university labs.

• ​​​​CTOs and engineering leaders from various companies.

• ​​​​VCs with a keen interest in the deep-tech domain.